DWMelon wrote:
123outerme wrote:
DWMelon wrote:
You do not need to put an appvar into ram for it to be read. This means you can pretty much have unlimited tilemaps. The only RAM that it will take up is the string that holds the tilemap data, which I believe is approximately 900 bytes. This should be a huge fix for your memory problems.
When I call det(0, the Appvar is temporarily loaded into RAM by Doors CSE, and it has to be, as far as I can tell. Otherwise, it is in Archive. It stops when the Appvar is loaded on top of the other 14k+ bytes of the main program. And if I'm wrong about this, someone who knows this for sure, correct me!
hmm... Well I asked KermM in SAX, and he said it was read directly from archive... Also if you look at tifreak's Lib Help program you will see that the appvar for it is over 40kb. I'm not sure how DCSE could load that into ram.
Anyways would be awesome if someone who knows for sure could comment and explain what is true.
Ok, thanks. It's probably because I have to store the 600 byte tilemap data twice then, once when it is read into Str9 and then into the string reserved for holding map data, Str1. Although that shouldn't be enough to crash the game, I guess it might be.
123outerme wrote:
DWMelon wrote:
You do not need to put an appvar into ram for it to be read. This means you can pretty much have unlimited tilemaps. The only RAM that it will take up is the string that holds the tilemap data, which I believe is approximately 900 bytes. This should be a huge fix for your memory problems.
When I call det(0, the Appvar is temporarily loaded into RAM by Doors CSE, and it has to be, as far as I can tell. Otherwise, it is in Archive. It stops when the Appvar is loaded on top of the other 14k+ bytes of the main program. And if I'm wrong about this, someone who knows this for sure, correct me!
I know in asm (and axe) you dan read directly from archive (but it will be slower). I would presume it would stay the same for color calcs. So if you cant, that is something that should be added to celtic
c4ooo wrote:
123outerme wrote:
DWMelon wrote:
You do not need to put an appvar into ram for it to be read. This means you can pretty much have unlimited tilemaps. The only RAM that it will take up is the string that holds the tilemap data, which I believe is approximately 900 bytes. This should be a huge fix for your memory problems.
When I call det(0, the Appvar is temporarily loaded into RAM by Doors CSE, and it has to be, as far as I can tell. Otherwise, it is in Archive. It stops when the Appvar is loaded on top of the other 14k+ bytes of the main program. And if I'm wrong about this, someone who knows this for sure, correct me!
I know in asm (and axe) you dan read directly from archive (but it will be slower). I would presume it would stay the same for color calcs. So if you cant, that is something that should be added to celtic
Maybe Kerm didn't implement it because it would be slower than just copying the data to RAM and reading from there. It's fine though, the solution isn't much worse than the program I had before I had these errors.
Quote:
When I call det(0, the Appvar is temporarily loaded into RAM by Doors CSE, and it has to be, as far as I can tell.
That's not correct. Doors CSE reads AppVars directly from Archive. Are you correctly accounting for the space that your program itself takes while it's running?
KermMartian wrote:
Quote:
When I call det(0, the Appvar is temporarily loaded into RAM by Doors CSE, and it has to be, as far as I can tell.
That's not correct. Doors CSE reads AppVars directly from Archive. Are you correctly accounting for the space that your program itself takes while it's running? I am. It has to be the loading process then. All the variables I load to start the game, plus two copies of the map data (once into Str9 with det(0 and then into Str1) and that's probably enough to crash it. Thanks for clarification Kerm!
Warning: Discussion of beta features, maps, etc. may include SPOILERS. Proceed at your own caution.
- Signing Up
If you would like to sign up for the closed beta, please PM me to avoid cluttering the topic. I won't accept many testers, to avoid many cooks in one kitchen, as well as to contain possible spoilers. If I accept you, you will be PM'd a link to the newest beta whenever they are released, as well as a save file editing program, and instructions.
As you will be able to modify your save file and experience a lot of content in a short time, I ask that if you would want to experience the game free from beta duties, that you not sign up. Other than that, go for it!
- Beta Information
As for what beta testers will be testing, I say just playtesting. If you see a possible bug, try to replicate it, and report it. If it is unreplicatable, report it anyways but make sure you specify that it's unreplicatable. Worlds 1 and 2 have been thoroughly tested. World 3 has been tested a bit, and World 4 hasn't been tested at all.
- Known Issues
*Not really an issue but I have to implement DJ Omnimaga's battle background code.
I hope you've gotten a lot of PMs asking to beta-test. I'd certainly be interested, but I'll send you a PM about the same. I'm not sure what else to comment, other than to congratulate you for getting to a beta phase. If you have more screenshots and information about the current state to share with us, perhaps that would help drive activity in the topic?
KermMartian wrote:
I hope you've gotten a lot of PMs asking to beta-test. I'd certainly be interested, but I'll send you a PM about the same. I'm not sure what else to comment, other than to congratulate you for getting to a beta phase. If you have more screenshots and information about the current state to share with us, perhaps that would help drive activity in the topic?
Thanks Kerm! I've gotten a few. I can make some screenshots.
I'm starting up work on this again. It's been on and off but that's been due to several games I've suddenly found more fun than I ever did, life stuff, and feeling burnout on this project. I have another idea for another project, but this one WILL be finished as well.
What you see here is the newest screenshot, with many features suggested by the testers who work with me. Not many tiles, if any, have changed since the last screenshot. Most of what changed is internal work, such as a small chance that you can attack first, even if you don't outspeed the enemy (and the opposite is also true). This isn't demonstrated in the screenshot, but it's an example.
Wow, I can't believe how far this has come! The sprites look spiffy, the interface is clean and cohesive, and I love that you can even rebind what the arrow keys do in battle. This will definitely be one of if not the standout RPG for xLIBC, and I can't wait for tr1p1ea and I to make it available on the TI-84 Plus CE too.
Keep up the great work.
KermMartian wrote:
Wow, I can't believe how far this has come! The sprites look spiffy, the interface is clean and cohesive, and I love that you can even rebind what the arrow keys do in battle. This will definitely be one of if not the standout RPG for xLIBC, and I can't wait for tr1p1ea and I to make it available on the TI-84 Plus CE too.
Keep up the great work.
Thanks so much! Actually, the rebinding thing you see is your different attacks. Each attack corresponds to an arrow key. However, you are somewhat right, you can set any attack to any arrow key. For example, you can have Slice mapped to Up (the first attack slot) and another attack mapped to Right (the last slot), with no attacks in any other slots. But other functions like Block, Run, and choosing your attack are set to specific keys.
And now, the pressure is on to really make this one shine. I'll do my best
Haha, I'm sorry, I didn't mean to put extra pressure on you. Ah, thanks for clarifying about the key mapping; that makes sense. I assume you already have all the beta-testers you want?
KermMartian wrote:
Haha, I'm sorry, I didn't mean to put extra pressure on you. Ah, thanks for clarifying about the key mapping; that makes sense. I assume you already have all the beta-testers you want?
No problem, it's less of pressure stressing me out, and more of pressure making me not lazy and work on it some more. I have a few testers, but more would always be appreciated! Especially if some testers get really busy or something. I assume this is an application, right?
Here's a new gif of weaknesses and upgraded attack animations! The yellow flash indicates that you've used a move that the enemy is weak to. Instead of just guesswork, the NPCs will also hint to/tell you the enemies' weaknesses. On top of this, NPCs will also offer other helpful advice (especially in the first dungeon) and offer to upgrade your attacks. (Although that last one was always present)
Having that hint seems like it will make gameplay more intuitive to me. Also, I like when games provide tutorial-like information as you play the early game without having an explicit tutorial for you to read or follow (or even a specific tutorial level). Kudos on that. And yes, that was indeed a beta-testing application from me.
Especially with certain tr1p1eas getting xLIBCE functions working in our private builds of Doors CE!
KermMartian wrote:
Having that hint seems like it will make gameplay more intuitive to me. Also, I like when games provide tutorial-like information as you play the early game without having an explicit tutorial for you to read or follow (or even a specific tutorial level). Kudos on that. And yes, that was indeed a beta-testing application from me.
Especially with certain tr1p1eas getting xLIBCE functions working in our private builds of Doors CE!
Okay, then I'll send you a build! Feel free to test it with xLIBCE, and if you can, send me some screenshots of how awesomely fast it (hopefully) is!
Shh... I didn't drop this preview of the fifth world here...
Pretty big update:
Thanks to some xLIBC feature and tr1p1ea's help, I've been able to speed up the saving process a ton! Unfortunately for testers, I had to slightly change the way some data was saved. I'll tell you guys how to fix that in the PM, once I send it all out. Here's the difference. I tried to be exact when I pressed the Save button, but it may not be perfect.
Wow, that's about twice the speed (half the time, if you prefer)! Good thing you came up with the optimization when the game was still in a private beta phase, so it'll impact relatively few users.